In This Issue

Immunohistochemistry (IHC) is commonly used in both laboratory and clinical settings to visualize tissue markers. Formalin fixation and paraffin embedding of tissues poses a challenge in IHC, because it can destroy epitopes recognized by the Abs. Furthermore, multiplex IHC requires primary Abs raised in different species to prevent cross-reactivity, thus limiting the choice of primary Abs. In this issue, Gorris et al. (p. 347) developed and optimized an eight-color multiplex IHC panel that includes five different immune checkpoint molecules (PD-1, PD-L1, OX40, CD27, and TIM3), CD3, a tumor marker, and DAPI. The approach took advantage of tyramide signal amplification (TSA), which is used to study the expression of multiple markers in a single tissue section and is more sensitive than conventional fluorescence or chromogenic IHC, enabling detection of low-abundance targets in high resolution. Because successful IHC using TSA hinges on the order in which the Abs are applied, the authors developed the panel sequentially, comparing multiplex staining to reference single stains, and determined that the order of primary Abs within the multiplex panel affects signal intensity and can underestimate or overestimate the signal, necessitating stepwise optimization of the order in which primary Abs are used. As a proof of principle, the multiplex panel was tested in different tumor tissues, using tissue microarrays consisting of invasive margin cores of both hematologic and solid tumors. These studies identified differences in immune checkpoint expression, not only among different tumors, but also within the same tumor sample. The described method, therefore, allows analysis of multiple immune checkpoints simultaneously within the tumor microenvironment and has applications in the clinic for selection of patients who are good candidates for checkpoint immunotherapy and for tracking their progress during treatment.

Epigenetic modifiers, such as the ten-eleven translocation (TET) family of methylcytosine dioxygenases, regulate T cell differentiation. Whereas TET2 has been shown to play a role in CD4⁺ Th cell differentiation, the function of TET2 in CD8⁺ T cell differentiation is unknown. Utilizing TET2^fl/flCD4Cre⁺ (TET2 conditional knockout [TET2cKO]) mice, Carty et al. (p. 82) elucidated the role of TET2 in CD8⁺ T cell memory differentiation. TET2cKO mice acutely infected with lymphocytic choriomeningitis virus (LCMV) had similar absolute numbers of splenic LCMV-specific CD8⁺ T cells, but lower frequencies of these cells in peripheral blood. Moreover, TET2 loss enhanced CD8⁺ T cell effector function, as evidenced by increased levels of intracellular IFN-γ and CD107a in TET2cKO compared with wild-type (WT) LCMV-specific CD8⁺ T cells. Infected mice displayed an increase in LCMV-specific memory precursor effector cells (MPECs), a decrease in short-lived effector cells (SLECs), and an increase in frequency and absolute number of central memory CD8⁺ T cells, suggesting that TET2 loss directs an early decision toward a central memory phenotype. When equal numbers of WT or TET2-deficient LCMV-specific P14 CD8⁺ T cells were transferred into congenic hosts that were subsequently infected with LCMV, a higher frequency of TET2-deficient P14 cells differentiated into MPECs, whereas WT P14 cells differentiated into SLECs. These results suggest that TET2 regulates CD8⁺ T cell differentiation in a cell-intrinsic manner. In a rechallenge model, WT congenic mice received gp33⁺ CD8⁺ T cells from TET2cKO LCMV-challenged donors and were subsequently infected with L. monocytogenes expressing the LCMV gp33 epitope. These mice exhibited greater pathogen control than those receiving CD8⁺ T cells from WT mice previously infected with LCMV. Finally, examination of genome-wide differentially methylated regions (DMRs) in CD8⁺ T cells from LCMV-infected TET2cKO mice revealed that Tbx21, PRDM1, IRF4 and Runx3 contained DMRs, suggesting that TET2 may regulate drivers of fate decisions, rather than exerting its function via direct demethylation of memory marker loci. Taken together, these data reveal a novel role of TET2 in regulation of CD8⁺ T cell effector versus memory cell fate decisions in an LCMV infection model.

Human leukocyte antigen-E, a highly conserved human nonclassical MHC-Ib molecule that binds to CD94/NKG2 receptors on NK cells, can also bind and present self- and pathogen-derived peptides to CD8⁺ T cells. Determining the impact of MHC-E–restricted CD8⁺ T cell responses on pathogen control requires nonhuman primate (NHP) models that mirror human MHC-E–restricted T cell biology. Wu et al. (p. 49) investigated MHC-E immunobiology in two macaque species commonly used in biomedical research. Whereas only two HLA-E alleles are found in humans, the MHC-E locus in rhesus macaques (RM) encodes at least 33 MHC-E alleles (Mamu-E), corresponding to 30 distinct molecules, and gives rise to a greater diversity of MHC-E alleles at both the population and the individual animal level. Conversely, Mauritian-origin cynomolgus macaques (MCM) express one or two alleles (Mafa-E) of limited diversity, closely resembling the diversity in humans at both the population and individual level. Despite these differences, MHC-E was similarly expressed on the surface of human, RM, and MCM monocytes, as well as T, B, and NK cells, and bound identical peptides, as evidenced by the successful binding by all three MHC-E varieties of four previously described SIVgag-derived Mamu-E peptide epitopes. Consistent with this, in vitro and in vivo studies showed that SIV-specific Mamu-E–restricted RM CD8⁺ T cells recognized peptides presented by all MHC-E varieties, indicating an ability to recognize Ag across heterologous APCs. In vitro infection of human, MCM, and RM CD4⁺ T cells with SIV led to increased MHC-E expression on all cell types, indicating that HLA-E, Mamu-E, and Mafa-E are similarly regulated by SIV infection. In vivo infection of RM and MCM with SIV showed similar upregulation of MHC-E on CD4⁺ T cells, but enhanced MHC-E expression was also observed on monocytes, NK, B, and CD8⁺ T cells, indicating a global response to viral infection. Thus, this study describes two NHP models that may be used to study MHC-E immunobiology and the role MHC-E–restricted CD8⁺ T cells may play in pathogen control. Given the limited MHC-E diversity across the population and the finding that MHC-E presents Ag to cells of both the innate and adaptive systems, MHC-E represents a novel therapeutic target for the development of CD8⁺ T cell-based vaccines.

Macrophages play critical roles both in initiating type 1diabetes (T1D) and combating Coxsackievirus infection, a suspected trigger of T1D. Recent studies have demonstrated the importance of oxidative stress in T1D, as deficiency in NADPH oxidase (NOX), which is expressed in macrophages at high levels, delays the onset of T1D in NOD mice by skewing macrophages from an M1 to an M2 phenotype and diminishes TLR3-dependent inflammatory responses. In this issue, Burg et al. (p. 61) sought to provide a mechanistic link between oxidative stress during diabetogenic viral infection and T1D. Coxsackievirus B3 (CB3)-infected NOX-deficient NOD mice (NOD.Ncf1^m1J) were significantly protected against virus-induced autoimmune T1D when compared with infected NOD mice, and this was associated with significant decreases in circulating CXCL10, CCL5, TNF-α, and IFN-β by 7 d postinfection. The observed delay in T1D development in CB3-infected NOD.Ncf1^m1J mice was also associated with reduced expression of antiviral genes in the pancreas. Recruitment of macrophages to the pancreas was similar between CB3-infected NOD and NOD.Ncf1^m1J mice, but macrophages recruited in the latter mice were less inflammatory, as TNF-α⁺ macrophages in these mice were reduced in percentage and absolute number and produced less TNF-α compared with those in CB3-infected NOD mice. Consistent with this, NOD.Ncf1^m1J macrophages infected in vitro with CB3 displayed reductions in Tnf, Cxcl10, Ccl5, Ifnb1, and Isg15 expression, as well as reduced expression of the viral RNA sensors Tlr3 and Ddx58. Importantly, addition of exogenous superoxide to NOD.Ncf1^m1J CB3-infected macrophages increased the levels of MDA5, a viral dsRNA sensor, to levels similar to those in infected NOD controls and rescued the antiviral M1 response in CB3-infected NOD.Ncf1^m1J, as indicated by increased expression of STAT1, a transcription factor involved in inflammatory M1 differentiation and type I IFN signaling. Collectively, these data demonstrate that innate immune signaling by macrophages following CB3 infection is reduction-oxidation regulated and provides a mechanistic link between diabetogenic viruses, oxidative stress, and autoimmunity.